Counting the Extensions of the Directory

[Mago]

Category:

Miscellaneous

Description:

Counting the Extensions of the Archives in Current Directory.

#!/usr/bin/perl -w
use strict;

my %extensions;

while (<*>) {
    if (-f && /\.(\W+)$/) {
        $extensions{$1}++;
    }
}

foreach $key (keys %extensions) {
    print "$key -> $extensions{$key}";
}

__END__

LINE:

perl -e 'while(<*>){if(-f && /\.(\W+)$/){$e{$1}++}}print "$_ -> $e{$_}
+\n" for keys %e'

[merlyn]

I'm not sure what an extension is, because it's apparently some file with a period followed by non-alphanums at the end of the name. That's pretty rare. Did you mean \w+ instead of \W+?

[Mago]

Ok ! Yes, you are certain, with \w+ is very better.



perl -e 'while(<*>){if(-f && /\.(\w+)$/){$e{$1}++}}print "$_ -> $e{$_}
+\n" for keys %e'

Thanks;

[EdwardG]

E:\tmp>ls
1gfJaJTeoh
CFG1584.tmp
CFG1587.tmp
EScan
ExchangePerflog_8484fa31f2d96a21cfcccd43.dat
Perflib_Perfdata_784.dat
SCP1D28.tmp
SCPA62.tmp
VBE
WINWORD.log
add_service_authentication_columns.sql
add_service_authentication_columns.sql.bak
ch01.pdf
create_export_tables-2.sql
create_export_tables-2.sql.bak
en-only.rws.swp
jusched.log
msohtml
msohtml1
perldoc_BDeparse_T439ed4fb_d50b1.txt
perldoc_perlfunc_T439eb5b6_bc83c.txt
plugtmp
svn000.tmp.pl
tmp158D.exe
tmp158D.tmp
tmp158E.tmp
tmp1592.tmp
tmp180E.tmp
tmp1811.tmp
~DF67F.tmp
~DF68C.tmp
~DF723C.tmp
~DF850E.tmp
~DFDDFC.tmp
~WRD0828.doc
~WRF2409.tmp
~WRS0003.tmp
~WRS0004.tmp

E:\tmp>ls | grep -oE "\.\w+$" | sort -u
.bak
.dat
.doc
.exe
.log
.pdf
.pl
.sql
.swp
.tmp
.txt
.xls

E:\tmp>

(GNU grep 2.5.1)

[Mago]

You are not counting the extensions ?



IBIZA desentol /tarifa1/users/desentol/mlima/dados/input > ls | grep -
+oE "\.\w+$" | sort -u
grep: illegal option -- o
usage: grep [-E | -F] [-c|-l|-q] [-insvxbhwy] [-p parasep] -e pattern_
+list...
        [-f pattern_file...] [file...]
usage: grep [-E | -F] [-c|-l|-q] [-insvxbhwy] [-p parasep] [-e pattern
+_list...]
        -f pattern_file... [file...]
usage: grep [-E | -F] [-c|-l|-q] [-insvxbhwy] [-p parasep] pattern [fi
+le...]

[EdwardG]

GNU grep 2.5.1 has the -o option to output only the matched text. Your grep doesn't.

As for counting...

E:\tmp>ls | grep -oE "\.\w+$" | sort -u | wc -l
     12